Scoring app using Python
In this article we are presenting how to develop the scoring agent using python and serialized model, developed with Jupiter notebook.
We are using the job stories to define the requirements:
- when a Reefer container telemetry event arrives, I want the scoring app to compute an anomaly detection predictive score so that it can create a reefer container maintance command event.
- when a Reefer container telemetry event arrives, I want the scoring app I want the data to be transform so the scoring can be done using expected structure.
- when the scoring app is deploy to kubernetes, I want to be sure it is healthy so that the kubernetes scheduler does not kill it
Create the project with Appsody
For this code we are using the same approach as for the simulator app development
Integrate test driven development
Deployent to openshift
- appsody deploy